Podcast | Identifying experience gaps, and applying service design thinking to improve CX


Director, Strategy & Content | Podcast Host of Mobile Interactions Now

20 min podcast
Frank Spillers

In this Part 2 of our conversation with Frank Spillers, the founder and CEO at Experienced Dynamics, we delve deeper into how to identify experience gaps, learn from usability researches and apply service design thinking to create better customer experience.

Podcast transcript


Frank, welcome back to the show. In our previous episode, we touched on some of the overall challenges that UX designers and service designers have to work with together with their business teams. In this episode, I would like to delve deeper into how to get better at applying some of the things we are learning from the new type of interactions between businesses and consumers, many of them triggered by the pandemic. What worked and what didn't and what we should be doing more of?



Well, thanks for having me back on the show, Jean. It's nice to be back. To ask the question of how well are we doing our current CX, which is a marketing capacity UX, which is usually a product focused capacity or service design, which is a holistic way to develop both service and product and digital experiences and CX in every channel together. How do you know? And then from a COVID perspective, what do you do as a starting point? What's the way to figure out if you're missing any opportunities in terms of your post-COVID experience?

To me, the answer to those questions is you've got to find out. It's like where we leave with measurement, that idea of measurement. You've got to always be listening, ear on the ground, so-called customer listening posts. We do this anyway with analytics, with sentiment analysis, we do it online with social media. No marketer would be caught dead without having engagement or analytics behind a social media campaign.

So, with UX and service design, we need those qualitative, those probes, if you will. We need constant user testing. We need service testing. We need to be able to give customers diaries over time and track their movements across channels over a two week, three week period, and measure how well ... And this is even without an existing solution. I mentioned that you have to do that once you actually design things right. Find the right problem, create a service blueprint if you're doing the service, and then measure it and roll it out and so forth, test it and so forth.

But you can start there as well. So, you can start with, okay, let's measure, what's our existing experience like? Where are the break points? Where are the broken touch points? Where's the channel falling down? You interview the customer and for an example they go, "I don't feel comfortable. When I deal with their online support people, they were really rude, or I didn't understand them or they didn't understand me, or I've never had a good experience. I want to go talk to someone in person."

Well, that's not just a case of, let's make sure that they have a good experience once they get to the store or to the organization, to the building. But why did the customer feel uncomfortable in the other channel? So, backing up and looking at their holistic. And then what were they trying to do? Why couldn't they just do this online? We offer an online way to do it.



When you are talking about this, it just remind me of one of the case studies from your work, actually, Wells Fargo contact center. And I found it really interesting how you guys noticed certain segment of customers are happy. The business people were being served very well and the consumer part of it wasn't. Can you tell us a little bit about how to even think about it, maybe that the segments are the problem, or how to approach what type of study you applied to solve these problems?

Yeah. The type of study is we use the same, which is behavioral research, which is ethnography, or it was chairside visits. The Wells Fargo, they call it phone banking, it's basically customer support for small businesses. So, small businesses call in. And what we literally did is just sat beside the call center people and plugged into their live calls and sat and listened to them all day long for an entire week. And so that's an example of this behavioral research where there is no focus group, there's no survey, you're not asking them for their opinions. That's market research, those two techniques. So, you want to do this realistic day in the life, walk in their shoes, chairside, whatever you want to call it, that comes from the discipline of anthropology and it's used in UX. Ethnography is a regular UX technique. Think of it as people watching, but calculated people watching.

So, in that story with the Wells Fargo, the thing you're talking about is there was almost VIP. So, if there's a problem, it gets pushed to a manager or a team of managers. There are various degrees. But the customers don't know that. It's a little bit like IT, how they have five degrees of escalation and you don't know where you are, so you're talking to tier one and someone at tier four can handle it, someone at tier five can just snap their fingers and it's gone, the problem. But your muddling your way through tier one, and then tier two, tier three or tier two, and then you get an angry and you have to ask for a manager.

So tier three comes in. That manager doesn't help you. This happened with Rackspace, actually, Rackspace built their company on Sensational, I think fanatical support. They service mark that, right? Fanatical support. And they had it and it was good and it was true and it was a shining light and then they were bought by another company and that customer support tanked it.

So, it's possible also to lose your traction. And that's why it's so important when you do this stuff. It's not just about getting excited about journey mapping or customer centricity. It's about managing all this in a program. And also, this is required if you're doing inclusive design, which is making sure underrepresented users are brought into the process as well. But the funny story to end the Wells Fargo piece is that ... I was actually a customer of Wells Fargo. So, I was taking my own personal notes about which, and I actually met one of the guys that deals with that VIP, that white glove service, which I didn't know about. And I took his business card and I went back to my company and gave it to my CFO and said, "If you ever need any help, this is the guy to call. Trust me."



So, we all have this anecdotal thing where when it works, it's really beautiful. As a business how do you do that more systematically? In a more scalable way that this permeates to more customers and longer time? And I think this is where the whole design thinking kicks in. Is there a way to do this better?



You see, the reality is most UX designers are fighting in their organization. Service designers are having to justify this weird, "Okay, I don't want to be stuck by your scope." Can imagine telling that to somebody, a product manager or to a engineering director? like, "I don't want to be stuck by your scope. I need to look at the whole thing holistically." They'd be like, "No. we need to know. What are we developing, an app? Is it a mobile app? How many screens does it have?" So, it requires a lot of education, but also organizational participation, like inclusive design. Requires that everyone be involved and everyone be doing it. Otherwise, you can just sprinkle a little bit of accessibility or can't sprinkle a little bit of inclusiveness DEI, just a DEI project over here, and then the rest of all your channels are excluding users or rubbing them the wrong way or committing bias and so forth and so on.

So, requires that systematic approach. And yes, it is totally possible. But I wanted to touch on something you said about the AI approach. The thing you laid out is a classic AI approach. Which is take someone who's performing really well, like maybe those people in the call center, and model them in terms of using them as a training set for an algorithm. I would recommend against that approach. The reason I recommend against that approach from a UX perspective is a story of a system we are redesigning at Nike. We were redesigning the system and the system had a total of six users. An internal system, total of six users. And this is a system that's used to track when athletes wear the swish. It's an internal system. It's managing this important thing, which is done through AI and cameras or whatever.

This one woman basically designed the system. Then we went and talked to the other users. One of the users, the person had a cheat sheet. It was directions that the super-user had written. The reason it doesn't work from a UX perspective is this cheat sheet that this person had, they never looked at it. So, that's user two of six or whatever, that person. And then we noticed it with the next person. They're like, "This system is so difficult. I have some training notes here. We were given a training." And the training notes are right there beside the desk, but they're not being read. They're not being used. And it's like, what's that all about? So, if you model the approach of that high-performing area of your company or channel…



Oh, you mean they are two different types.



Yeah, you might lock other people out and then they can't figure it out.



The other day, I was talking with a Microsoft Dynamics integrator. They were using our Conversations API connector, because what they're doing is… it's a same day trading scenario where the relationship between the agent and the customer is very important because the transaction has to close on the same day. What they are learning, to my surprise, is that although their first interaction is bot generated, the human agent has to actually send it, because the customers are so loving the private messages they've been getting every day. The template-based, bot based communication needs to have a human carrier to send to them because that is what the end user is expecting. If it's an important enough customer that needs a dedicated agent to begin with, that is. And you just start embedding some of the things, the templates and links there, that can lead to more bot-enabled interactions, should they choose it.



Conversational UIs as well as speech UIs, speech recognition, are extremely difficult to get right. In other words, they require a lot of very, very important UX work to be done. And there's a lot of known things and limitations about those systems and so forth and so on. The people that have made them useful after many, many years and much money, like Google and Amazon and so forth, did a lot of things and also went through a lot of dead space, if you will, where for example, the Amazon Echo didn't understand accents until a couple of years ago, even though it's been around for ...



I'm still not happy with it.



Yeah. Right. Yeah, exactly. It's for an American male Silicon valley accent, American white middle-class male accent. It's even got bias where it'll recognize an English accent, but not a Scottish or Irish accent, or Australian accent, but an English accent is okay. Anyway, to say that conversation and speech require a lot of very, very important work. Same with bots. Bot interaction can just be miserable. I had an experience with Amazon last Christmas. I happened to be buying a laptop, not for Christmas, just was replacing a staff laptop in kind of an emergency, but not that big of a deal. Except it was during Amazon's rush period for Apple laptops. And instead of telling the truth about that ... So, Apple said, "Need three weeks or four weeks or six weeks to deliver." Apple said that. Amazon said, "Can get it to you in two days."

So I ordered it and then it disappeared from my orders. And I was like, "That's weird. I thought I did that. Maybe I was distracted. I'll do it again." So, I ordered it again. It stayed in my orders for about 10 hours and then it disappeared. No trace of it. And I'm like, "I thought I ordered that. I'm sure I ordered it at least once. Maybe I don't see it. I really need this. Let me just get that ordered." Ordered it a third time. And now I'm four or five days into the, "I should've had it already." The same thing again, disappeared.

So, I contacted support and it had the laptop. The only record of it was in the chat bot. I thought, "Aha! I did order it."






"Something's really sinister here." So, I selected that and they connected me to an agent. That agent put me to a second agent. That agent put me to a third agent. That agent put me to a fourth agent.

At that point I asked for the manager and all I was asking was, why? Because there was no money lost. Amazon just basically lied and they couldn't handle the Christmas fulfillment. I went through, I think, seven different operators. And even the manager was not able to handle me and just passed me on to somebody else who was a lower level. And I thought, my God, what a complete disaster of a chat bot experience.

So, most people are not doing chat bots well, because chat bot design is a whole discipline within UX. You have to do it right. And yes, you're right, it does point to the problem of AI modeling expertise, or actually trying to get automation to take over when it really should be a human hand. And so understanding where you have the opportunity to put a human hand, the machine's hands, human, machine, human, it's going to be a blended future. We're not going to just wake up and you order your latte by speech, you look at the wall and it pays for the latte. You think a thought and it creates a 3D hologram that brings your friends together and you hang out. It's not going to be as futuristic as that, but it'll probably be more augmented back and forth.



I actually look forward to that progression, knowing that some things can go towards more of that direction. And this is my question for you. What do you want to see more of happening?



Yeah. I've learned not to take my own experiences too personally, like some other experts in the field, just sort of bitch about technology. So to me, the beautiful thing is not a fixed technology. It would be nice if more and more organizations understood they also had this other modality, which is service design, that they could use to holistically think about redesigning their whole ecosystem and the product service experience within that.

So, that would be one aspect. I think the bigger one relates to the book I'm working on, which is UX management. So, in other words, let's say you're like, "Okay, Frank, we heard you about service design. We're doing the damn blueprint and we're going to go through the process." And so you hire service designers or some of these clients that I worked for years ago who are now hiring service design directors, you hire service designers, you figure it out, you integrate it, you launch a product and/or a service or whatever, you fix it, you repair it, whatever.

It comes out. Oh, that was good. That was useful. They're seeing some value with that. How do you do it again? And let's say then at successful, it goes out into the marketplace. It mitigates a bunch of risks. It adds a lot of value. How do you do it again? Then you start hiring more team, you're growing, you hire another 10, 15, 20 UX type folks, you even hire a few service designers. How do you get that entire team to work with the rest of the company? Doesn't understand a word of what they're saying, for one thing. It's like, "What do you mean touch points and cross-channel interactions, or orchestration, or real-time interaction management as a new category of analytics that's come out, what do you mean by all that stuff? And what does it..."

So, to me, what I'd like to see, is more organizations truly collaborating without silos and doing UX as a team sport in a way that actually helps improve collaboration and helps get a better business outcome for both the organization and the end user customer.



So, I lied a little bit. I do have one little nosy question for you. Can you tell us, so that we get to know you better, what are the three things you do most on your phone?



Well, I try not to use my phone too much and that's a very deliberate strategy because I'm trying to conserve my attention, which is a IT worker's 21st century dilemma. But the apps that I use a lot are Seconds, is the name of the app. It's a custom timer app, and you can create your own timer. So, I use that for various, various things. The other app that I'm on all the time is YouTube or an app that removes the ads called Tube Browser, which is all right. That's not a plug for that because it sometimes takes a long time, but it does work eventually. But YouTube, I listen to a lot of stuff on my commute. And then I think probably the third most heavily used app would be probably, I want to say WhatsApp or Messenger or Outlook, something like that.

But I use Google Translate a lot. I also use, because I work in different languages or speak different languages as well, so I'm using that all the time to continue that learning. But the third one, I would leave the category open and say I'm always looking for a new app that solves a particular problem, fills a particular need. For example, I just downloaded an app that's called Birthdays. And what I want to do is put all the birthdays and take that off of Facebook because it's one of Facebook's hidden features, is they keep track of the birthdays and it's a user adoption, you go, "Oh." You don't remember, but it remembers for you. Even though it's nothing to do with the experience, it's actually a core emotional, social ... So, I'm actually trying to create my own shadow app for that because I'm not on Facebook and I don't use Facebook at all anymore since the last two years, I think. So yeah, the last category is any apps that helps do something that either improves either a productivity or a personal wellness or a social need.



Sounds good. I think you are looking for something useful. And on that note, I really, really thank you for the conversation… and good luck with your books and we should touch base after the book is released and off to press.



Sounds good. Thank you, Jean.